CN115085196B - Power load predicted value determination method, device, equipment and computer readable medium - Google Patents

Abstract

Embodiments of the present disclosure disclose a method, device, device and computer-readable medium for determining a power load forecast value. A specific implementation of the method includes: obtaining a historical power load data set; performing clustering processing on the historical power load data in the historical power load data set to obtain a historical power load data class set; using the historical power load data class set, from Determine the key influencing factors in the preset initial influencing factor set to obtain the key influencing factor set; generate the target decision tree according to the historical power load data class set and the key influencing factor set; input the key influencing factor vector of the date to be predicted into the target decision tree , to get the classification result; use the classification result to determine the power load forecast value of the date to be predicted. This embodiment can improve the accuracy of the finally generated electric load forecast value, and enhance the guiding significance of the electric load forecast value to the electric power system.

Description

Method, device, device, and computer-readable medium for determining electric load forecast value

technical field

The embodiments of the present disclosure relate to the field of computer technology, and in particular to a method, device, device and computer-readable medium for determining a power load forecast value.

Background technique

Power load value forecasting is a technology that uses past historical power load data to predict the power load value of the power system at a certain time in the future. At present, when forecasting the power load value, the usual way is to use the time series method, regression analysis method or cluster data mining technology to predict the power load value.

However, when using the above method to predict the power load value, there are often the following technical problems:

First, when the time series method or regression analysis method is used, the prediction accuracy is low, which will affect the stability of the power supply system;

Second, when clustering data mining technology is used, too many parameters are required, and when the amount of data is large, the prediction takes a long time and the efficiency is low.

Contents of the invention

The Summary of the Disclosure is provided to introduce concepts in a simplified form that are described in detail in the Detailed Description that follows. The content of this disclosure is not intended to identify the key features or essential features of the claimed technical solution, nor is it intended to be used to limit the scope of the claimed technical solution.

Some embodiments of the present disclosure provide a method, device, device and computer readable medium for determining a power load forecast value, so as to solve one or more of the technical problems mentioned in the background art section above.

In the first aspect, some embodiments of the present disclosure provide a method for determining a power load forecast value. The method includes: acquiring a historical power load data set, wherein the historical power load data in the historical power load data set is used to represent a certain The power load curve of a date; the historical power load data in the above-mentioned historical power load data set is clustered to obtain the historical power load data class set; using the above-mentioned historical power load data class Determine the key influencing factors in , and obtain the set of key influencing factors; according to the above-mentioned historical power load data set and the above-mentioned key influencing factor set, a target decision tree is generated, wherein, the nodes in the above-mentioned target decision tree are respectively related to the above-mentioned historical power load data set Corresponding to the historical power load data class in ; input the key influencing factor vector of the date to be predicted into the above-mentioned target decision tree, and obtain the classification result; use the above-mentioned classification result to determine the power load forecast value of the above-mentioned date to be predicted.

Optionally, generating a target decision tree according to the set of historical power load data and the set of key influencing factors includes:

Using the hierarchical clustering tree as an initial decision tree, perform the following steps for each node in the initial decision tree in turn:

In response to determining that there are two leaf nodes at the node, determining two historical power load data classes corresponding to the two leaf nodes in the historical power load data class set as historical power load data class pairs;

Using the set of key influencing factors, determine the key influencing factor with the largest difference between the two historical power load data classes in the historical power load data class pair;

Generating classification rules for the nodes by using the key influencing factors with the largest differences;

Combining the initial decision tree and each classification rule of the nodes in the initial decision tree to obtain the target decision tree.

Optionally, using the set of key influencing factors to determine the key influencing factors with the greatest difference between the two historical power load data classes in the pair of historical power load data classes includes:

Determining the average value of each key influencing factor in each historical power load data class in the historical power load data class pair, wherein the average value of each key influencing factor is the same as that in the set of key influencing factors One-to-one correspondence of key influencing factors;

Determining the difference between the average values of the two key influencing factors corresponding to the same key influencing factor among the average values of the key influencing factors of the two historical electric load data categories in the historical power load data class pair, to obtain the set of mean value differences;

The key influencing factor corresponding to the average value difference with the largest value in the set of average value differences is determined as the key influencing factor with the largest difference.

Optionally, the generating the classification rule of the node by using the key influencing factor with the largest difference includes:

The average value of the average values of the two key influencing factors corresponding to the key influencing factor with the largest difference is determined as the classification value;

A classification rule for the node is generated by using the classification value.

Optionally, before using the classification result to determine the predicted value of the electric load on the date to be predicted, the method further includes:

The initial power load forecasting model is trained by using each historical power load data in the classification result to obtain a target power load forecasting model.

Optionally, using the classification result to determine the predicted value of the electric load on the date to be predicted includes:

Using the target power load forecasting model to generate a power load forecast value for each power user to obtain a set of power load forecast values;

The power load prediction values in the power load prediction value set are summed to obtain the power load prediction value.

In the second aspect, some embodiments of the present disclosure provide a device for determining a predicted value of electric power load. The device includes: an acquisition unit configured to obtain a historical power load data set, wherein the historical power load in the historical power load data set The data is used to represent the power load curve of a certain date; the clustering unit is configured to perform clustering processing on the historical power load data in the above-mentioned historical power load data set to obtain a historical power load data set; the first determination unit, It is configured to use the above-mentioned set of historical power load data to determine the key influencing factors from the preset initial set of influencing factors to obtain a set of key influencing factors; the generating unit is configured to The factor set generates a target decision tree, wherein the nodes in the above target decision tree correspond to the historical power load data classes in the above historical power load data class set respectively; the input unit is configured to convert the key influencing factor vector of the date to be predicted The above-mentioned target decision tree is input to obtain a classification result; the second determining unit is configured to use the above-mentioned classification result to determine the electric load forecast value of the above-mentioned date to be predicted.

In a third aspect, some embodiments of the present disclosure provide an electronic device, including: one or more processors; The processor executes, so that one or more processors implement the method described in any implementation manner of the first aspect above.

In a fourth aspect, some embodiments of the present disclosure provide a computer-readable medium on which a computer program is stored, wherein when the program is executed by a processor, the method described in any implementation manner of the above-mentioned first aspect is implemented.

The above-mentioned various embodiments of the present disclosure have the following beneficial effects: the electric load forecast value obtained through the method for determining the electric load forecast value in some embodiments of the present disclosure has higher accuracy. Specifically, the reason for the low accuracy of the power load forecast value obtained by the relevant power load value forecasting method is that the time series method or regression analysis method only needs a small amount of historical data to complete the power load forecast value, and it is difficult to comprehensively and fully Estimate characteristics of historical electrical load data. Based on this, the method for determining the predicted value of electric load in some embodiments of the present disclosure first obtains a set of historical electric load data, wherein the historical electric load data in the above-mentioned historical electric load data set is used to represent the electric load curve of a certain date . Thus, according to actual application requirements, a set of historical power load data in a long period of time can be obtained. Then, cluster processing is performed on the historical power load data in the above-mentioned historical power load data set to obtain a historical power load data cluster. Therefore, historical power load data with relatively similar characteristics can be grouped into one group, which is convenient for subsequent prediction. Then, using the above-mentioned set of historical power load data, the key influencing factors are determined from the preset initial influencing factor set to obtain the key influencing factor set. Thus, according to the clustering results of the historical power load data, the key influencing factors that have a greater impact on the clustering results can be determined from the initial influencing factor set. Next, generate a target decision tree according to the above-mentioned set of historical power load data and the set of key influencing factors, wherein the nodes in the above-mentioned target decision tree correspond to the historical power load data in the above-mentioned set of historical power load data respectively. Thus, the target decision tree can be generated by combining the clustering results and key influencing factors. Then, input the key influencing factor vector of the date to be predicted into the above-mentioned target decision tree to obtain the classification result. Finally, by using the above classification results, the electric load prediction value of the above date to be predicted is determined. Thus, each historical power load data in the historical power load data set can be fully and effectively utilized. Therefore, the accuracy of the final generated power load forecast value is improved, the guiding significance of the power load forecast value to the power system is enhanced, and the dynamic balance of the power system load and the stability of the entire power system are guaranteed to a certain extent.

Description of drawings

The above and other features, advantages and aspects of the various embodiments of the present disclosure will become more apparent with reference to the following detailed description in conjunction with the accompanying drawings. Throughout the drawings, the same or similar reference numerals denote the same or similar elements. It should be understood that the drawings are schematic and elements and elements have not necessarily been drawn to scale.

FIG. 1 is a flow chart of some embodiments of a method for determining a power load forecast value according to the present disclosure;

Fig. 2 is a structural schematic diagram of some embodiments of the device for determining the predicted value of electric load of the present disclosure;

FIG. 3 is a schematic structural diagram of an electronic device suitable for implementing some embodiments of the present disclosure.

detailed description

Embodiments of the present disclosure will be described in more detail below with reference to the accompanying drawings. Although certain embodiments of the disclosure are shown in the drawings, it should be understood that the disclosure may be embodied in various forms and should not be construed as limited to the embodiments set forth herein. Rather, these examples are provided so that the understanding of this disclosure will be thorough and complete. It should be understood that the drawings and embodiments of the present disclosure are for exemplary purposes only, and are not intended to limit the protection scope of the present disclosure.

It should also be noted that, for the convenience of description, only the parts related to the related invention are shown in the drawings. In the case of no conflict, the embodiments in the present disclosure and the features in the embodiments can be combined with each other.

It should be noted that concepts such as "first" and "second" mentioned in this disclosure are only used to distinguish different devices, modules or units, and are not used to limit the sequence of functions performed by these devices, modules or units or interdependence.

It should be noted that the modifications of "one" and "multiple" mentioned in the present disclosure are illustrative and not restrictive, and those skilled in the art should understand that unless the context clearly indicates otherwise, it should be understood as "one or more" multiple".

The names of messages or information exchanged between multiple devices in the embodiments of the present disclosure are used for illustrative purposes only, and are not used to limit the scope of these messages or information.

The present disclosure will be described in detail below with reference to the accompanying drawings and embodiments.

Referring to FIG. 1 , FIG. 1 shows a process 100 of some embodiments of a method for determining an electric load forecast value according to the present disclosure. The method for determining the electric load forecast value includes the following steps:

Step 101, acquiring a historical power load data set.

In some embodiments, the subject of execution of the method for determining the predicted value of electric load may obtain the historical electric load data set through a wired connection or a wireless connection. Wherein, the historical power load data in the above historical power load data set can be used to characterize the power load curve of a certain date. The above historical power load data set may correspond to the power load curve of each day in a historical time period.

As an example, the time span of the above historical time period may be 1 year.

Step 102, clustering the historical power load data in the historical power load data set to obtain the historical power load data set.

In some embodiments, the execution subject clusters the historical power load data in the historical power load data set to obtain the historical power load data set, which may include the following steps:

The first step is to perform hierarchical clustering on the historical power load data in the above historical power load data set to obtain a hierarchical clustering tree. Wherein, the agglomerative hierarchical clustering algorithm can be used to cluster the historical power load data in the above historical power load data set.

The second step is to determine each historical power load data at each node in the hierarchical clustering tree as a historical power load data class to obtain a set of historical power load data classes.

In step 103, the key influencing factors are determined from the pre-set initial influencing factor set by using the historical power load data set, and the key influencing factor set is obtained.

In some embodiments, the initial influencing factors in the aforementioned set of initial influencing factors may include but not limited to at least one of the following: maximum temperature, minimum temperature, average temperature, average humidity, wind speed and precipitation. The above-mentioned executive body uses the above-mentioned historical power load data class set to determine the key influencing factors from the preset initial influencing factor set, and obtain the key influencing factor set, which may include the following steps:

The gray correlation analysis is carried out on the above-mentioned historical power load data set and the above-mentioned initial influencing factor set to obtain the key influencing factor set.

As an example, the key influencing factors in the above set of key influencing factors may include average temperature and precipitation.

Step 104, generating a target decision tree according to the historical power load data set and the key influencing factors set.

In some embodiments, the executive body may use the CART (Classification And Regression Tree, Classification and Regression Tree) algorithm to generate a target decision tree according to the above-mentioned historical power load data class set and the above-mentioned key influencing factor set.

The nodes in the above-mentioned target decision tree respectively correspond to the historical electric load data classes in the above-mentioned historical electric load data class collection. The above-mentioned executive body can generate a target decision tree according to the above-mentioned historical power load data set and the above-mentioned key influencing factor set through the following steps:

In the first step, the above-mentioned hierarchical clustering tree is used as the initial decision tree, and the following steps are performed for each node in the above-mentioned initial decision tree in turn:

In the first sub-step, in response to determining that there are two leaf nodes in the node, determine the two historical power load data classes corresponding to the two leaf nodes in the historical power load data class set as historical power load data class pairs.

The second sub-step is to determine the key influencing factor with the greatest difference between the two historical power load data classes in the above historical power load data class pair by using the above key influencing factor set.

The third sub-step is to use the above-mentioned key influencing factors with the largest differences to generate the classification rules for the above-mentioned nodes.

In the second step, the above-mentioned initial decision tree and each classification rule of the nodes in the above-mentioned initial decision tree are combined to obtain the above-mentioned target decision tree.

In some optional implementations of some embodiments, the execution subject uses the set of key influencing factors to determine the key influencing factor with the largest difference between the two historical power load data classes in the historical power load data class pair, which may include The following steps:

The first step is to determine the average value of each key influencing factor of each historical power load data in each historical power load data class in the above-mentioned historical power load data class pair. Wherein, the above-mentioned average value of each key influencing factor corresponds to the key influencing factors in the above-mentioned key influencing factor set one by one.

The second step is to determine the difference between the average values of the two key influencing factors corresponding to the same key influencing factor among the average values of the key influencing factors of the two historical power loading data categories in the above-mentioned historical power load data class pair, and obtain the mean difference collection of values. Wherein, the average value difference in the above average value difference set is an absolute value.

The third step is to determine the key influencing factor corresponding to the average value difference with the largest value in the above average value difference set as the key influencing factor with the largest difference.

Optionally, the above-mentioned executive body uses the above-mentioned key influencing factors with the largest difference to generate the classification rules of the above-mentioned nodes, which may include the following steps:

In the first step, the average value of the average values of the two key influencing factors corresponding to the above-mentioned key influencing factors with the largest difference is determined as the classification value.

In the second step, the classification rules of the above nodes are generated by using the above classification values. Wherein, the classification rule may include two sub-classification rules, which are used to divide a group of numeric values into two groups. The two sub-classification rules may be respectively that the data to be classified is greater than the above classification value, and the data to be classified is less than or equal to the above classification value.

The above steps, as an inventive point of the embodiments of the present disclosure, solve the technical problem 2 mentioned in the background technology, "When clustering data mining technology is used, too many parameters are required, and when the amount of data is large, the prediction consumes The time is longer and the efficiency is lower." The factors that lead to the above technical problems are often as follows: Clustering data mining algorithms, such as CURE (Clustering Using Representatives) algorithm, have many parameters, sampling errors, sensitivity to differences in spatial data density, and low prediction efficiency when the amount of data is large. If the above factors are solved, the effect of improving the forecasting efficiency can be achieved. In order to achieve this effect, the present disclosure uses the above-mentioned hierarchical clustering tree as an initial decision tree, and performs the following steps for each node in the above-mentioned initial decision tree in turn: In response to determining that there are two leaf nodes in the above-mentioned node, the above-mentioned historical power The two historical power load data classes corresponding to the above two leaf nodes in the load data class set are determined as historical power load data class pairs. Using the above set of key influencing factors, determine the key influencing factor with the largest difference between the two historical power load data classes in the above historical power load data class pair. The above-mentioned key influencing factors with the largest differences are used to generate the classification rules for the above-mentioned nodes. Therefore, the target decision tree can be generated by using the results of hierarchical clustering and the set of key influencing factors, so that the preliminary forecast results can be quickly obtained only by using the forecast date, and the forecasting efficiency of power load data can be greatly improved.

Step 105, input the key influencing factor vector of the date to be predicted into the target decision tree to obtain the classification result.

In some embodiments, the above key influencing factor vector includes an actual value of the key influencing factor corresponding to each key influencing factor in the above key influencing factor set. The above-mentioned executive body inputs the key influencing factor vector of the date to be predicted into the above-mentioned target decision tree, and obtains the classification result, which may include the following steps:

In the first step, according to the actual value of the key influencing factor corresponding to each key influencing factor in the above key influencing factor set included in the above key influencing factor vector, determine the historical power corresponding to the data of the above-mentioned date to be predicted in the above-mentioned target decision tree Load data class. The values of each dimension in the key influencing factor vector above are the actual values corresponding to each key influencing factor.

The second step is to determine the corresponding historical power load data class as the above classification result.

Step 106, using the classification result to determine the predicted value of electric load on the date to be predicted.

In some embodiments, the execution subject uses the above classification result to determine the electric load forecast value of the date to be predicted, which may include the following steps:

In the first step, the target power load forecasting model is used to generate the power load forecast value of each power user, and a set of power load forecast values is obtained.

The second step is to sum the power load prediction values in the above power load prediction value set to obtain the power load prediction value.

In some optional implementation manners of some embodiments, the above execution subject may also send the above electric load forecast value to the target terminal for display. Wherein, the above-mentioned target terminal can be used to display the above-mentioned power load forecast value for reference by decision makers.

Optionally, before using the classification result to determine the power load forecast value of the date to be predicted, the execution subject can also use the historical power load data in the classification result to train the initial power load forecasting model to obtain the target Electric load forecasting model. Wherein, the above initial power load forecasting model may be a CNN (Convolutional Neural Networks, Convolutional Neural Networks) model.

The above steps, as an inventive point of the embodiments of the present disclosure, further solve the technical problem 2 mentioned in the background technology "When using clustering data mining technology, too many parameters are required, and when the amount of data is large, the prediction consumes The time is longer and the efficiency is lower.” The factors that lead to the above technical problems are often as follows: Clustering data mining algorithms, such as CURE (Clustering Using Representatives) algorithm, have many parameters, sampling errors, sensitivity to differences in spatial data density, and low prediction efficiency when the amount of data is large. If the above factors are solved, the effect of improving the forecasting efficiency can be achieved. In order to achieve this effect, the present disclosure first uses each historical power load data in the above classification results to train an initial power load forecasting model to obtain a target power load forecasting model. Then, the target power load forecasting model is used to generate the power load forecast value of each power user, and a set of power load forecast values is obtained. Finally, the power load prediction values in the above power load prediction value set are summed to obtain the power load prediction value. Thus, a preliminary prediction can be made on the electric load value on the date to be predicted according to the above-mentioned target decision tree. Then, the initial power load forecasting model is trained by using each historical power load data in the classification result. Finally, use the trained target power load forecasting model to make accurate forecasts. While improving efficiency, accuracy is ensured.

The above-mentioned various embodiments of the present disclosure have the following beneficial effects: the electric load forecast value obtained through the method for determining the electric load forecast value in some embodiments of the present disclosure has higher accuracy. Specifically, the reason for the low accuracy of the power load forecast value obtained by the relevant power load value forecasting method is that the time series method or regression analysis method only needs a small amount of historical data to complete the power load forecast value, and it is difficult to comprehensively and fully Estimate characteristics of historical electrical load data. Based on this, the method for determining the predicted value of electric load in some embodiments of the present disclosure first obtains a set of historical electric load data, wherein the historical electric load data in the above-mentioned historical electric load data set is used to represent the electric load curve of a certain date . Thus, according to actual application requirements, a set of historical power load data in a long period of time can be obtained. Then, cluster processing is performed on the historical power load data in the above-mentioned historical power load data set to obtain a historical power load data cluster. Therefore, historical power load data with relatively similar characteristics can be grouped into one group, which is convenient for subsequent prediction. Then, by using the above-mentioned historical power load data set, the key influencing factors are determined from the preset initial influencing factor set to obtain the key influencing factor set. Thus, according to the clustering results of the historical power load data, the key influencing factors that have a greater impact on the clustering results can be determined from the initial influencing factor set. Next, generate a target decision tree according to the above-mentioned set of historical power load data and the set of key influencing factors, wherein the nodes in the above-mentioned target decision tree correspond to the historical power load data in the above-mentioned set of historical power load data respectively. Thus, the target decision tree can be generated by combining the clustering results and key influencing factors. Then, input the key influencing factor vector of the date to be predicted into the above-mentioned target decision tree to obtain the classification result. Finally, by using the above classification results, the electric load prediction value of the above date to be predicted is determined. Thus, each historical power load data in the historical power load data set can be fully and effectively utilized. Therefore, the accuracy of the final generated power load forecast value is improved, the guiding significance of the power load forecast value to the power system is enhanced, and the dynamic balance of the power system load and the stability of the entire power system are guaranteed to a certain extent.

Further referring to FIG. 2 , as an implementation of the methods shown in the above figures, the present disclosure provides some embodiments of a device for determining a power load forecast value, and these device embodiments correspond to those method embodiments shown in FIG. 1 , The device can be specifically applied to various electronic devices.

As shown in FIG. 2 , an apparatus 200 for determining a predicted electric load value in some embodiments includes: an acquisition unit 201 , a clustering unit 202 , a first determination unit 203 , a generation unit 204 , an input unit 205 and a second determination unit 206 . Wherein, the acquisition unit 201 is configured to acquire a historical power load data set, wherein the historical power load data in the above historical power load data set is used to characterize the power load curve of a certain date; the clustering unit 202 is configured to The historical power load data in the above-mentioned historical power load data set is clustered to obtain the historical power load data set; Determine the key influencing factors in the set to obtain the set of key influencing factors; the generation unit 204 is configured to generate a target decision tree according to the above-mentioned historical power load data class set and the above-mentioned key influencing factor set, wherein the nodes in the above-mentioned target decision tree are respectively Corresponding to the historical power load data class in the above-mentioned historical power load data class collection; the input unit 205 is configured to input the key influencing factor vector of the date to be predicted into the above-mentioned target decision tree to obtain the classification result; the second determination unit 206 is It is configured to use the above classification result to determine the electric load forecast value of the date to be predicted.

It can be understood that the units recorded in the device 200 correspond to the steps in the method described with reference to FIG. 1 . Therefore, the operations, features and beneficial effects described above for the method are also applicable to the device 200 and the units contained therein, and will not be repeated here.

Referring now to FIG. 3 , it shows a schematic structural diagram of an electronic device 300 suitable for implementing some embodiments of the present disclosure. The electronic device shown in FIG. 3 is only an example, and should not limit the functions and scope of use of the embodiments of the present disclosure.

As shown in FIG. 3 , an electronic device 300 may include a processing device (such as a central processing unit, a graphics processing unit, etc.) 301 that can be randomly accessed according to a program stored in a read-only memory (ROM) 302 or loaded from a storage device 308 Various appropriate actions and processes are executed by programs in the memory (RAM) 303 . In the RAM 303, various programs and data necessary for the operation of the electronic device 300 are also stored. The processing device 301 , ROM 302 and RAM 303 are connected to each other through a bus 304 . An input/output (I/O) interface 305 is also connected to the bus 304 .

Typically, the following devices can be connected to the I/O interface 305: input devices 306 including, for example, a touch screen, touchpad, keyboard, mouse, camera, microphone, accelerometer, gyroscope, etc.; including, for example, a liquid crystal display (LCD), speaker, vibration output device 307 such as a device; and a communication device 309. The communication means 309 may allow the electronic device 300 to perform wireless or wired communication with other devices to exchange data. While FIG. 3 shows electronic device 300 having various means, it should be understood that implementing or having all of the means shown is not a requirement. More or fewer means may alternatively be implemented or provided. Each block shown in FIG. 3 may represent one device, or may represent multiple devices as required.

In particular, according to some embodiments of the present disclosure, the processes described above with reference to the flowcharts may be implemented as computer software programs. For example, some embodiments of the present disclosure include a computer program product, which includes a computer program carried on a computer-readable medium, where the computer program includes program codes for executing the methods shown in the flowcharts. In some such embodiments, the computer program may be downloaded and installed from a network via communication means 609 , or from storage means 608 , or from ROM 602 . When the computer program is executed by the processing device 601, the above functions defined in the methods of some embodiments of the present disclosure are performed.

It should be noted that the computer-readable medium described in some embodiments of the present disclosure may be a computer-readable signal medium or a computer-readable storage medium or any combination of the above two. A computer readable storage medium may be, for example, but not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any combination thereof. More specific examples of computer-readable storage media may include, but are not limited to, electrical connections with one or more wires, portable computer diskettes, hard disks, random access memory (RAM), read-only memory (ROM), erasable Programmable read-only memory (EPROM or flash memory), optical fiber, portable compact disk read-only memory (CD-ROM), optical storage device, magnetic storage device, or any suitable combination of the above. In some embodiments of the present disclosure, a computer-readable storage medium may be any tangible medium that contains or stores a program that can be used by or in conjunction with an instruction execution system, apparatus, or device. In some embodiments of the present disclosure, however, a computer-readable signal medium may include a data signal propagated in baseband or as part of a carrier wave, carrying computer-readable program code therein. Such propagated data signals may take many forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination of the foregoing. A computer-readable signal medium may also be any computer-readable medium other than a computer-readable storage medium, which can transmit, propagate, or transmit a program for use by or in conjunction with an instruction execution system, apparatus, or device . Program code embodied on a computer readable medium may be transmitted by any appropriate medium, including but not limited to: wires, optical cables, RF (radio frequency), etc., or any suitable combination of the above.

In some embodiments, the client and the server can communicate using any currently known or future-developed network protocols such as HTTP (HyperText Transfer Protocol), and can communicate with digital data in any form or medium (for example, communication networks) interconnections. Examples of communication networks include local area networks (“LANs”), wide area networks (“WANs”), internetworks (e.g., the Internet), and peer-to-peer networks (e.g., ad hoc peer-to-peer networks), as well as any currently known or future developed network of.

The above-mentioned computer-readable medium may be included in the above-mentioned electronic device, or may exist independently without being incorporated into the electronic device. The above-mentioned computer-readable medium carries one or more programs, and when the above-mentioned one or more programs are executed by the electronic device, the electronic device: acquires a historical power load data set, wherein the historical power load data set in the above-mentioned historical power load data set The power load data is used to represent the power load curve of a certain date; the historical power load data in the above-mentioned historical power load data set is clustered to obtain the historical power load data class set; using the above-mentioned historical power load data class set, from The key influencing factors are determined in the preset initial influencing factor set to obtain the key influencing factor set; according to the above-mentioned historical power load data set and the above-mentioned key influencing factor set, a target decision tree is generated, wherein the nodes in the above-mentioned target decision tree are respectively related to Corresponding to the historical power load data class in the above historical power load data class collection; input the key influencing factor vector of the date to be predicted into the above target decision tree to obtain the classification result; use the above classification result to determine the power load forecast value of the date to be predicted .

Computer program code for carrying out operations of some embodiments of the present disclosure may be written in one or more programming languages, or combinations thereof, including object-oriented programming languages—such as Java, Smalltalk, C++, Also included are conventional procedural programming languages - such as the "C" language or similar programming languages. The program code may execute entirely on the user's computer, partly on the user's computer, as a stand-alone software package, partly on the user's computer and partly on a remote computer or entirely on the remote computer or server. In cases involving a remote computer, the remote computer can be connected to the user computer through any kind of network, including a local area network (LAN) or a wide area network (WAN), or it can be connected to an external computer (for example, using an Internet service provider to connected via the Internet).

The flowchart and block diagrams in the Figures illustrate the architecture, functionality, and operation of possible implementations of systems, methods and computer program products according to various embodiments of the present disclosure. In this regard, each block in a flowchart or block diagram may represent a module, program segment, or portion of code that contains one or more logical functions for implementing specified executable instructions. It should also be noted that, in some alternative implementations, the functions noted in the block may occur out of the order noted in the figures. For example, two blocks shown in succession may, in fact, be executed substantially concurrently, or they may sometimes be executed in the reverse order, depending upon the functionality involved. It should also be noted that each block of the block diagrams and/or flowchart illustrations, and combinations of blocks in the block diagrams and/or flowchart illustrations, can be implemented by a dedicated hardware-based system that performs the specified functions or operations , or may be implemented by a combination of dedicated hardware and computer instructions.

The units described in some embodiments of the present disclosure may be realized by software or by hardware. The described units may also be set in a processor. For example, it may be described as: a processor includes an acquisition unit, a clustering unit, a first determination unit, a generation unit, an input unit, and a second determination unit. Wherein, the names of these units do not constitute a limitation to the unit itself under certain circumstances, for example, the acquisition unit may also be described as "a unit for acquiring historical power load data sets".

The functions described herein above may be performed at least in part by one or more hardware logic components. For example, without limitation, exemplary types of hardware logic components that may be used include: Field Programmable Gate Arrays (FPGAs), Application Specific Integrated Circuits (ASICs), Application Specific Standard Products (ASSPs), System on Chips (SOCs), Complex Programmable Logical device (CPLD) and so on.

Claims (8)

1. A method for determining a power load forecast value, comprising: Obtaining a historical power load data set, wherein the historical power load data in the historical power load data set is used to represent the power load curve of a certain date; performing clustering processing on the historical power load data in the historical power load data set to obtain a historical power load data set; Using the set of historical power load data to determine key influencing factors from a preset set of initial influencing factors to obtain a set of key influencing factors; Generate a target decision tree according to the set of historical power load data classes and the set of key influencing factors, wherein the nodes in the target decision tree correspond to the historical power load data classes in the historical power load data class set respectively , the generating the target decision tree includes: using a hierarchical clustering tree as an initial decision tree, performing the following steps for each node in the initial decision tree in turn: In response to determining that there are two leaf nodes at the node, determining two historical power load data classes corresponding to the two leaf nodes in the historical power load data class set as historical power load data class pairs; Determining the average value of each key influencing factor in each historical power load data class in the historical power load data class pair, wherein the average value of each key influencing factor is the same as that in the set of key influencing factors One-to-one correspondence of key influencing factors; determine the difference between the average values of two key influencing factors corresponding to the same key influencing factor in the average values of each key influencing factors of the two historical electric load data classes in the historical power load data class, A set of mean difference values is obtained, wherein the mean difference value in the set of mean difference values is an absolute value; the key influencing factor corresponding to the mean difference value with the largest numerical value in the set of mean difference values is determined as the difference The biggest key influencers; Using the key influencing factor with the largest difference to generate the classification rule of the node includes: determining the average value of the average values of the two key influencing factors corresponding to the key influencing factor with the largest difference as a classification value; using the classification value Generating classification rules for the nodes, wherein the classification rules include two sub-classification rules for dividing a group of numerical values into two groups, the two sub-classification rules are respectively that the data to be classified is greater than the classification value, and the data to be classified is The data is less than or equal to the classification value; combining the initial decision tree and each classification rule of the nodes in the initial decision tree to obtain the target decision tree; Input the key influencing factor vector of the date to be predicted into the target decision tree to obtain the classification result; Using each historical power load data in the classification results to train the initial power load forecasting model to obtain the target power load forecasting model, wherein the initial power load forecasting model is a CNN (Convolutional Neural Networks, convolutional neural network) model ; Using the classification result, determining the predicted power load value of the date to be predicted includes: using the target power load forecast model to generate the predicted power load value of each power user, and obtaining a set of predicted power load values; The power load prediction values in the power load prediction value set are summed to obtain the power load prediction value. 2. The method according to claim 1, wherein said clustering the historical power load data in the historical power load data set to obtain the historical power load data class set includes: performing hierarchical clustering on the historical electric load data in the historical electric load data set to obtain a hierarchical clustering tree; Each historical power load data at each node in the hierarchical clustering tree is determined as a historical power load data class to obtain a set of historical power load data classes. 3. The method according to claim 1, wherein the initial influencing factors in the initial influencing factor set include but are not limited to: maximum air temperature, minimum air temperature, average air temperature, average humidity, wind speed and precipitation; and Using the set of historical power load data to determine key influencing factors from a preset set of initial influencing factors to obtain a set of key influencing factors, including: A gray relational analysis is performed on the set of historical power load data and the set of initial influencing factors to obtain a set of key influencing factors. 4. The method according to claim 1, wherein the key influencing factor vector comprises an actual value of the key influencing factor corresponding to each key influencing factor in the set of key influencing factors; and The key influencing factor vector of the date to be predicted is input into the target decision tree to obtain classification results, including: According to the actual value of the key influencing factor corresponding to each key influencing factor in the key influencing factor set included in the key influencing factor vector, determine the historical power corresponding to the data of the date to be predicted in the target decision tree load data class; The corresponding historical power load data category is determined as the classification result. 5. The method according to any one of claims 1-4, wherein the method further comprises: Sending the electric load forecast value to the target terminal for display. 6. A device for determining a power load forecast value, comprising: An acquisition unit configured to acquire a historical power load data set, wherein the historical power load data in the historical power load data set is used to represent a power load curve on a certain date; The clustering unit is configured to cluster the historical power load data in the historical power load data set to obtain a historical power load data set; The first determining unit is configured to use the set of historical power load data to determine key influencing factors from a preset set of initial influencing factors to obtain a set of key influencing factors; The generation unit is configured to generate a target decision tree according to the set of historical power load data types and the set of key influencing factors, wherein the nodes in the target decision tree are respectively related to the nodes in the set of historical power load data types Corresponding to the historical power load data class, the generation of the target decision tree includes: using the hierarchical clustering tree as the initial decision tree, and performing the following steps for each node in the initial decision tree in turn: In response to determining that there are two leaf nodes at the node, determining two historical power load data classes corresponding to the two leaf nodes in the historical power load data class set as historical power load data class pairs; Determining the average value of each key influencing factor in each historical power load data class in the historical power load data class pair, wherein the average value of each key influencing factor is the same as that in the set of key influencing factors One-to-one correspondence of key influencing factors; determine the difference between the average values of two key influencing factors corresponding to the same key influencing factor in the average values of each key influencing factors of the two historical electric load data classes in the historical power load data class, A set of mean difference values is obtained, wherein the mean difference value in the set of mean difference values is an absolute value; the key influencing factor corresponding to the mean difference value with the largest numerical value in the set of mean difference values is determined as the difference The biggest key influencers; Using the key influencing factor with the largest difference to generate the classification rule of the node includes: determining the average value of the average values of the two key influencing factors corresponding to the key influencing factor with the largest difference as a classification value; using the classification value Generating classification rules for the nodes, wherein the classification rules include two sub-classification rules for dividing a group of numerical values into two groups, the two sub-classification rules are respectively that the data to be classified is greater than the classification value, and the data to be classified is The data is less than or equal to the classification value; combining the initial decision tree and each classification rule of the nodes in the initial decision tree to obtain the target decision tree; The input unit is configured to input the key influencing factor vector of the date to be predicted into the target decision tree to obtain a classification result; The training unit is configured to use each historical power load data in the classification result to train the initial power load forecasting model to obtain a target power load forecasting model, wherein the initial power load forecasting model is CNN (Convolutional Neural Networks , convolutional neural network) model; The second determination unit is configured to use the classification result to determine the predicted power load value of the date to be predicted, including: using a target power load forecast model to generate a power load predicted value for each power user to obtain a power load predicted value gather; The power load prediction values in the power load prediction value set are summed to obtain the power load prediction value. 7. An electronic device comprising: one or more processors; a storage device on which one or more programs are stored, When the one or more programs are executed by the one or more processors, the one or more processors are made to implement the method according to any one of claims 1-5. 8. A computer-readable medium, on which a computer program is stored, wherein, when the program is executed by a processor, the method according to any one of claims 1-5 is implemented.

Images